Buying a process intelligence software startup may not seem a natural move for ABBYY, an established vendor of content IQ and document processing solutions. However, the company claims the purchase of TimelinePI is a natural extension of its successful drive to expand its integration with leading robotic process automation (RPA) platforms.

ABBYY Chief Sales Officer Bodo Wagener admits, “It may seem odd at the first glance, but if you look into the logic behind this deal, it’s actually a very smart move. ABBYY has positioned itself in the RPA space, and we are closely collaborating with the Blue Prism’s and UiPath’s of the world”. 

“TimelinePI is another nice complement to this market. The RPA industry is looking to improve processes and TimelinePI is actually the tool that helps the end user to understand where the problem actually is.  It provides a way to understand bottlenecks in the company and then RPA will follow. 

“One of the biggest challenges with RPA right now, is to identify the most suitable use cases. If you understand the processes, it’s much easier for you to find a suitable use case.  We believe that this TimelinePI acquisition is actually very complementary to ABBYY and our ecosystem of not only RPA vendors, but also BPM vendors and ECM vendors, in workflow and approval as well.”

Joseph Rayfield, EVP Sales and Business Development for TimelinePI, explains, “There seems to be two major concerns around truly expanding and accelerating our RPA initiatives. 

“The first one is, no one knows what to automate or why they’re automating it, so identifying, justifying, prioritising what to automate is critical if we’re going to make RPA a core enterprise function. 

“The second stumbling block, which becomes increasingly important, is whether everything is working as we expect it to work post implementation.  Believe it or not, these robotic processes don’t always run at maximum levels of efficiency.  So, understanding how they’re performing on a daily basis, and understanding how they affect the rest of the end-to-end process is important.

“It really stems from our desire to create an environment of sustainable and continuously improving automation. It is important to accept that if you are going to accelerate a number of steps by 100% or more, it is inevitably going to have an effect on the downstream process flow, and we also need to acknowledge that as processes change over time it will affect the robotic execution. So, staying on top of process operations on an ongoing basis is extremely important as we significantly expand out robotic deployments.

“Furthermore, being able to really interact with ‘in flow’ processes, to the point where you can detect a business condition like a missing credit check or missing fraud check, and being able to automatically spawn a digital worker to remedy that situation, starts creating ‘true’ closed loop remediation and significant additional value in automation”

Over the last 25 years, the cofounders of Timeline PI have created multiple enterprise software companies:  including data integration, business process management and a business intelligence platform, two of these subsequently being sold to webMethods and Kofax respectively.

“So, without giving their ages away, they’ve spent a long time involved in processes and analytics so they really understand this emerging space well,” said Rayfield.

“To me Timeline PI is the natural evolution of process analytics. You’ve got the manual approaches of Lean and Six Sigma which are really powerful, but suffer from being costly with regards to both time and money, and it is also, if we are honest, partially subjective.  We have also seen an influx of Process Mining, which can be very interesting but, to this day, tends to be more of a retrospective view of how you have previously performed rather than anything pushing towards operational process performance monitoring or predictive capabilities. 

“One of the most interesting thing about process analysis is that it has a dramatically reduced the time to value when compared to existing types of analysis. It really only needs three things too get started: a time stamp that something’s happened, an activity log of what happened and a unique identifier of who or what it happened to. 

“So, the data requirements for our type of analysis are just a fraction of what you’d need for BI.  And if you think of any modern system of record, they all have a capability to export those three specific fields.  Now obviously we like attributional data like who was involved in the process, which time zone was it in, which department was it, etcetera.

“But realistically, if I can get the three mandatory data points, I can successfully provide process analytics – automatically displaying sequencing of events, bottlenecks, missed steps, repeated steps and more…... 

“To help we’ve got multiple out of the box connectors like ServiceNow or Salesforce and it’s very simple to do real time integration with other leading platforms.  For some legacy systems, there’s a bit more work to do but what we’ve found that the data is generally available and it is certainly worth getting to. Finally, TimelinePI has an extensive ETL function that can help all forms of data preparation – to simplify and automate as much data manipulation as required.

“Realistically, Process Intelligence is fast becoming a strategic imperative – regardless of industry, all organizations have a myriad of hundreds, if not thousands of different processes, some of them very horizontally applicable with some vertically specific.  The TimelinePI vision is that in the future, all organisations will have libraries of ‘living’ interactive digital twins that will be continuously monitoring how their processes are executed, and how they are affecting the experience of the internal and external customer,” said Rayfield.